The amounts of data managed by companies are increasing sharply in line with the progress of the information society. Consequently, computer systems, which couple together host computers and storage apparatuses via switches and hubs, for example, computer systems, which make use of a SAN (Storage Area Network) to enable flexible data management, are widely used. The increase in companies specializing in the provision and management of IT systems to ordinary companies in recent years has resulted in the consolidation of data in the computer systems possessed by these specialized companies providing and managing IT systems. Thus, computer systems are becoming increasingly larger in size. This increase in the size of the computer system is also increasing the number of objects (for example, the number of volumes provided by a storage apparatus) handled by management software for managing these computer systems.
From the standpoint of managing a computer system, performance monitoring is important for the normal operation of the computer system. The management software is able to integratively manage the operating status and operational performance of resources comprising a computer system by collecting performance information from agents disposed in multiple management targets. The rise in the number of resources being managed is also increasing the amount of performance information to be acquired by the management software, thus requiring the processing of large amounts of performance information. PTL 1 discloses a method, which, based on information acquired by the management software, automatically adjusts the subsequent targeted range and acquisition frequency for acquiring performance information.